Artificial intelligence is becoming an increasingly familiar part of legal practice, with firms using AI tools for tasks ranging from drafting and document review to research, summarisation and administrative work.
The SRA’s new Warning Notice on Misuse of AI, published on 17 August 2026, is an important reminder that the use of AI does not change or reduce a solicitor’s existing professional obligations.
The SRA recognises that AI can provide significant benefits, but it has identified particular concerns about two areas: inaccurate or fabricated information produced by AI and the risks to client confidentiality when information is entered into AI systems.
AI does not replace professional responsibility
Perhaps the most important message from the Warning Notice is that solicitors remain responsible for the work they produce, regardless of whether AI has been used to prepare it.
AI-generated content must therefore be subject to appropriate human review and professional judgement. The fact that an AI system has produced an answer, citation, summary or piece of drafting does not make that output reliable or remove the solicitor’s responsibility for checking it.
This is particularly important in litigation. AI systems can generate convincing but entirely fictitious case names, citations, quotations and legal propositions. The SRA refers to a number of cases where inaccurate AI-generated material has found its way into court documents.
The message is straightforward: never assume that an AI-generated legal authority or factual assertion is correct. It must be independently checked and verified before it is relied upon.
Confidentiality is another major concern
The SRA is equally concerned about confidential client information being entered into AI tools.
Firms need to understand what happens to information submitted to an AI system, including whether it is stored, retained, accessed by third parties or used to train models. A paid subscription does not, by itself, mean that a system is suitable for processing confidential client information.
The Warning Notice emphasises that client information should only be entered into an AI system where appropriate contractual, technical and organisational safeguards are in place.
This is therefore not simply an IT issue. It potentially engages a firm’s obligations concerning client confidentiality, legal professional privilege, data protection and information security.
What should law firms be doing?
The SRA does not prohibit the use of AI. Instead, it expects firms to take a proportionate, risk-based approach.
Firms should consider whether they have:
- a clear policy governing the use of AI;
- identified which AI tools staff are permitted to use;
- assessed the risks associated with those tools;
- established rules about what information can and cannot be entered into AI systems;
- appropriate contractual and technical safeguards where client information is processed;
- arrangements for checking and verifying AI-generated content;
- appropriate supervision of AI-assisted work;
- provided staff with suitable training; and
- a clear process for reporting and dealing with errors, breaches or inappropriate use.
It is also important to remember that responsibility cannot simply be delegated to the technology. Managers and supervisors remain responsible for ensuring that work is appropriately supervised and that those using AI are competent to do so.
Don’t wait for an AI-related incident
For many smaller firms, the immediate priority does not need to be developing a complex AI governance framework. A sensible starting point is to establish some basic ground rules.
If your firm has not already done so, decide which AI tools may be used, for what purposes, by whom, and what information may be entered into them. Staff should also understand that publicly available AI tools should not be treated as secure repositories for confidential client information.
The SRA’s Warning Notice is a timely reminder that adopting AI is not simply a question of improving efficiency. Firms must ensure that innovation takes place within the existing framework of professional, regulatory and legal obligations.
Read the full SRA Warning Notice – Misuse of AI.
Legal Compliance Services can assist firms with reviewing their existing policies and procedures and developing a proportionate AI policy.
